مقاله
Job Scheduling Based on Single and Multi-Objective Meta-Heuristic Algorithms in Cloud Computing: A Survey
1394/12/12
کنفرانس،در International Conference on Information Technology, Communications and Telecommunications (IRICT)
Recently, cloud computing has attracted great attention as one of the most prominent technologies in the world. Companies like amazon, google, and microsoft are increasingly developing cloud-based systems by improving services to users. On the other hand, there are challenges in efficient use of cloud computing, including parallel processing, energy consumption, virtualizations and scheduling. The latter is considered as a major challenge. Scheduling determines how user requests are allocated to virtual machines in order to improve the quality of service. This study examined the existing task scheduling algorithms. First, the concepts of scheduling, type of layer, scheduling in various layers and main parameters for improving the quality of service are presented then, the most recent important meta-heuristic algorithms are analyzed.. These algorithms were designed as single objective- and multi-objective schedules to optimize different parameters in scheduling. Finally, evaluation methods and tools are presented.
همکاران :
نسیم سلطانی، بهرنگ برکتین، بهزاد سلیمانی نیسیانی،
Methods of Feature Extraction for Detecting the Duplicate Bug Reports in Software Triage Systems
1394/12/12
کنفرانس،در International Conference on Information Technology, Communications and Telecommunications (IRICT)
NOWADAYS, MANY USERS USE SOFTWARE REPOSITORIES SUCH AS QUESTION AND ANSWERING SITES, SOFTWARE TRIAGE SYSTEMS AND OTHER KIND OF REPOSITORIES THAT SAVE MANY TOPICS ABOUT THE SOFTWARE. ONE OF THE MOST IMPORTANT IN THIS AREA IS FINDING DUPLICATE ITEMS IN THESE REPOSITORIES ESPECIALLY FOR BUG REPORTS. IN THIS PAPER A REVIEW IS REPRESENTED IN THE STATE OF THE ART IN THIS RESEARCH AREA AND CATEGORIZE DIFFERENT KINDS OF FEATURES USED FOR FINDING DUPLICATES. ALSO THE METHODOLOGY OF RESEARCHES DESCRIBED THAT COULD BE USED AS A FRAMEWORK FOR FUTURE WORKS. FINALLY, IN CONCLUSION SECTION, CURRENT ISSUES IN THIS TOPIC MENTIONED AND GOALS AND IMPORTANCE OF THIS RESEARCH AREA WILL BE DESCRIBED.
همکاران :
بهزاد سلیمانی نیسیانی، سید مرتضی بابامیر،
Heuristic Algorithms for Task Scheduling in Cloud Computing: A Survey
1396/05/17
مجله،در International Journal of Computer Network and Information Security
Cloud computing became so important due to virtualization and IT systems in this decade. It has introduced as a distributed and heterogeneous computing pattern to sharing resources. Task Scheduling is necessary to make high performance heterogeneous computing. The optimization of related parameters, and using heuristic and meta-heuristic algorithms can lead to a reduction of the search space complexity and execution time. So, several studies have tried using a variety of algorithms to solve this issue and improve relative efficiency in their environments. This paper considered examines existing heuristic task scheduling algorithms. First, the concepts of scheduling, the layer of cloud computing, especially scheduling concept in the SaaS and PaaS layer, the main limits for improving the quality of service, evaluation methods of algorithms and applied tools for evaluating these ideas and practical experimental used methods were discussed and compared. Finally, future works in this area were also concluded and a summary of this article is presented in the form of a mind map.
همکاران :
نسیم سلطانی، بهزاد سلیمانی نیسیانی، بهرنگ برکتین،
Investigating Hadoop Architecture and Fault Tolerance in Map-Reduce
1396/04/09
مجله،در International Journal of Computer Science and Network Security
Map-Reduce is often used in implementation of critical and important tasks such as analysis of the scientific data. However, evidences in the past indicate the presence of optional errors that can destroy the results of Map-Reduce. Of course, run times of Map-Reduce like Hadoop can tolerate crash errors, but do not tolerate arbitrary or Byzantine errors. Hence in this paper, at first, the Hadoop architecture in distributed system will be investigated and then Hadoop will be compared with Map-Reduce and finally the Map-Reduce fault tolerance will be investigated.
همکاران :
آرمین کشکولی، بهزاد سلیمانی نیسیانی، مینا رهبری،
Automatic Verification of UML State Chart by BOGOR Model Checking Tool Automatic Formal Verification of Network and Distributed Systems
1394/08/15
کنفرانس،در International Conference on Knowledge-Based Engineering and Innovation (KBEI)،2
Validation and verification of software or system specifications are crucial in reducing costs and proper software development. Software specifications are usually represented by semi-formal languages like UML. For verification of non-formal and semi-formal models, they should be first transformed into a formal language. The state chart is one of the well-known UML charts that describe the behavior of a system and used for modeling many systems such as resource managements and communications in networks or distributed systems. In this paper, we propose a method to automatically map a UML state chart to BIR language, which is designed for BOGOR model checking. The goal of the verification in this paper is to evaluate the deadlock property of this chart. The proposed method is evaluated by four case studies of ATM and Fax machine state charts and the model is verified regarding the existence of a deadlock. Results indicate that while the PAT verification tool cannot properly recognize deadlocks in a state chart, the proposed approach is capable of detecting such cases of a deadlock. (3) Automatic Verification of UML State Chart by BOGOR Model Checking Tool Automatic Formal Verification of Network and Distributed Systems.
همکاران :
بهزاد سلیمانی نیسیانی، سید مرتضی بابامیر،
A New Algorithm for Mining Frequent Patterns in CAN tree CAN Growth Mining
1394/08/15
کنفرانس،در International Conference on Knowledge-Based Engineering and Innovation (KBEI)،2
Association Rule Mining is concerned with the search for relationships between item-sets based on co-occurrence of patterns. Since transactional databases are being updated all the time and there are always data being added or deleted, so Incremental Association Rule Mining is very important. Many methods have been presented so far for incremental frequent pattern mining, one of these methods is the frequent patterns mining base on the CAN tree (CANonical-order TREE). Related works on CAN tree, didn't discuss about extraction of frequent patterns from the tree and it has only been suggested that the mining method would be similar to FP-growth. In this paper, a new method is presented for mining CAN tree, and it is evaluated to show its improvement over the FP-growth method that mine FP tree. The evaluation results have demonstrated that performance of the presented algorithm is better than the FP-growth algorithm at high minimum support thresholds and for future work can try to improve it for lower minimum support threshold.
همکاران :
معصومه سادات حسینی، محمد ندیمی شهرکی، بهزاد سلیمانی نیسیانی،
Formal verification of UML statecharts using the LOTOS formal language
1394/08/15
کنفرانس،در International Conference on Knowledge-Based Engineering and Innovation (KBEI)،2
UML is a standard modeling language in software engineering. Although this language is capable of describing and modeling different aspects of a problem, it cannot be used for verification of the obtained model. Formal languages can be utilized for this purpose to verify the model. In a previous study by the authors of this paper, Statecharts diagram has been mapped to the LOTOS formal language. However, not all possible structures and necessary relations like the condition and loop structures were mapped. In this paper, an improvement of the previous method is presented. Moreover, for verification of the presented mapping, the CADP toolbox has been used. To apply the proposed method in practice, a case study is presented, where the properties deadlock, live lock, unreachable states, and non-deterministic states are formally verified. The results of the model verification show that the evaluated Statecharts diagram has had deadlocks, but there have not been live locks, unreachable states, or non-deterministic states in it.
همکاران :
محمد جوانی، بهزاد سلیمانی نیسیانی، سید مرتضی بابامیر،
A Framework for Improving Find Best Marketing Targets Using a Hybrid Genetic Algorithm and Neural Networks
1394/08/15
کنفرانس،در International Conference on Knowledge-Based Engineering and Innovation (KBEI)،2
Recently, many companies in Iran use telemarketing to introduce their products. These companies need to detect their best target to following them over seasons and years for more sales. This paper introduces a simple and appropriate method to predict behavior of customers based on behavior of prior customers. First of all, a dataset of customer action should be made and then preprocessed to reduce its attribute and dimension. Then a neural network will be made based on selected features to predict sale behavior of customers. Finally an evolutionary algorithm like genetic can be used to find feature of customers who will buy products more. This method evaluated by Portuguese Bank Tele Marketing dataset. Results show it simply can find the best customers in this case study. It's highly recommended to companies use this method to reduce their marketing costs and have better performance.
همکاران :
بهزاد سلیمانی نیسیانی، نسیم سلطانی، شیما قزلباش،
Recommendation Systems Based on Association Rule Mining for a Target Object by Evolutionary Algorithms
1397/02/18
مجله،در Emerging Science Journal،2
Recommender systems are designed for offering products to the potential customers. Collaborative Filtering is known as a common way in Recommender systems which offers recommendations made by similar users in the case of entering time and previous transactions. Low accuracy of suggestions due to a database is one of the main concerns about collaborative filtering recommender systems. In this field, numerous researches have been done using associative rules for recommendation systems to improve accuracy but runtime of rule-based recommendation systems is high and cannot be used in the real world. So, many researchers suggest using evolutionary algorithms for finding relative best rules at runtime very fast. The present study investigated the works done for producing associative rules with higher speed and quality. In the first step Apriori-based algorithm will be introduced which is used for recommendation systems and then the Particle Swarm Optimization algorithm will be described and the issues of these 2 work will be discussed. Studying this research could help to know the issues in this research field and produce suggestions which have higher speed and quality.
همکاران :
حسین حاتمی ورزنه، بهزاد سلیمانی نیسیانی، حسن ضیافت، نسیم سلطانی،